Helping the Poor in Latin America: Three Studies on Public Transfers, Social Services, and Poverty in Different Institutional Contexts

The goals of this dissertation are twofold. First, the dissertation aims to estimate and to explain the impact of public transfers and public provision of social services on poverty in Latin America. Second, the dissertation evaluates how the institutional context, embodied by the welfare state, shapes the impact of public transfers on poverty in Latin America. To achieve these goals, three studies compose the dissertation. The first study (Chapter 2) includes a classification of the policy orientations in Latin American countries. To classify policy orientations, an ideal typology was developed. While this first study of the dissertation finds that some countries fit more poorly than others, there is also some support for the ideal typology. With respect to the supportive evidence, the findings indicate that three clusters of countries fit the corporatist, universalist, and productivist ideal types of welfare states. The results also indicate the existence of two additional clusters that do not clearly fit in any of the ideal types of welfare states: a residual corporatist welfare model and a mix of corporatism and productivism. Finally, to illustrate the relevance of the welfare state classification, the first study of the dissertation also examined the connection between clusters and socioeconomic outcomes. The findings revealed that the classification of Latin American welfare states is particularly relevant for poverty and informal work. The second study (Chapter 3) intends to achieve the first goal of the dissertation by examining the impact of the redistributive budget allocated to benefits in cash and in kind on poverty levels among 15 Latin American countries in the period 1980–2000. With respect to social services, the findings of cross-country analysis indicate that the cumulative health and education spending does not have a significant impact on poverty rates. The cross-country analysis also included a measure of the education investment’s outcome, which were the years of education for the population aged 25 and older. The analysis reveals that this outcome measurement clearly contributed to reducing poverty in Latin America in the period 1980–2000. All in all, the findings of the dissertation suggest that investment in human capital does help to reduce poverty, but it is an open question if increasing human capital is associated with the intervention of the welfare state in Latin America. With respect to public transfers, the hypotheses of the second study suggested that the size of government budget allocated to public transfers will decrease poverty levels in pioneer corporatist and universalist welfare states, and the poverty reduction is higher in the latter. By contrast, with both clusters, another hypothesis says that the size of government budget allocated to public transfers would be associated with higher poverty levels in residual corporatist, productivist, and mixed Latin American welfare states. The empirical analysis verified these expectations. Finally, the third study (Chapter 4) is an examination of social assistance in Chile and evaluates the impact of social assistance transfers on chronic and transitory poverty with household panel data for Chilean working-age individuals in 2001 and 2006. It is argued that social assistance should affect only chronic poverty, increasing the levels of this type of poverty in Chile. The findings verified the prognosis for chronic poverty and suggested that some kind of indirect behavioral effect on labor supply results from the design of social assistance programs in Chile. As explanatory mechanisms, the study suggests that cash welfare is associated with incentives to apply for benefits and to work in the informal sector, and, in doing so, to increase chronic poverty. For transitory poverty, the hypothesis was, however, rejected. The results indicate also a positive effect of benefits on this type of poverty, but the magnitude and sensitivity of this finding are lower than the results for chronic poverty and thus suggest that further research is necessary to verify my findings